 |
 |
Archives of the TeradataForum
Message Posted: Thu, 02 Mar 2006 @ 14:43:18 GMT
| Subj: | | Table and join sequences.. |
| |
| From: | | Anomy Anom |
<-- Anonymously Posted: Thursday, March 02, 2006 06:09 -->
Dear all,
I ve come into TD with some Oracle experience. In writing Oracle queries, we follow a general rule to select the least bulkiest table last in
the SELECT statement. What is mean is If tabA and tabB are two tables having 1 and 10000 rows respectively, I would write a query like SELECT
count(*) from tabB, tabA than SELECT count(*) from tabA, tabB coz oracle works from 'right to left' . Same for the case of joins where we are
advised to put the most restrictive join at the end.
Any such rules necessary for teradata? If there are a series of equi-joins spawning a few intermediate spool files in the process, how do I
arrange the join sequence such that the query would consume the least amount of resources ?
Many regards
| |
